Comparison of GPS Receivers

Introduction

        GPS receivers have become a mainstay of forestry companies to accurately identify interesting locations. This means that on forest roads, culverts, bridges and washouts can be located for future reference and to fix any problems that occur. Since these items have a accurate location associated with the point, in the case of culverts and bridges when the road is removed the contractors do not have to hunt for the location. This is also helpful for situations when someone must return to the same location year after year, for example permanent sample plots. These are located some distance from a road or other landmark and must be relocated every couple of years to identify what has happened on the point. with the use of a GPS receiver it is possible to return to approximately the same location quicker. GPS is also used to layout clearcut harvesting blocks. this requires someone to walk through the forest and flag the boundary of the harvest area. this occurs under all types of forest cover and densities. As a class we preformed an experiment to compare the accuracy of a number of receivers under different forest conditions.

        The units that were used in this experiment were from Trimble, Garmin, and Sokkia. In total 6 units were used, 3 from Trimble, 2 Garmin's, and 1 Sokkia. The Trimble units were the Trimble XT, Trimble Geo 3, and the Trimble PRoXRS. The Garmin units that were used are the Garmin Map76 and the Garmin eTrex Legend. The only Sokkia unit that was used was the Sokkia Axis3. Some of the units have a number of ways of collecting information to correct the raw GPS data collected from the satellites including WAAS, raw data (Differentially corrected or Uncorrected), and the beacon. Table 1 shows how these data collection modes relate to each unit. Table 2 shows when the data was collected for each unit, it also shows how many times each unit was used to try and collect information. The unit was used till the first time it collected data for the point. For example the Garmin units colleted information on the first day and were not tried again.

        The whole system of using GPS receivers was created by the US military to allow for high precision locating of people and equipment. This is done using 24 satellites that are orbiting the world, generally in equatorial orbits. The location of the GPS receiver is determined through the distance the receiver is from a number of satellites, through a process called triangulation. the distance from a satellites is determined by the time a radio signal takes to reach the receiver. Each satellite transmits a slightly different signal allowing for differentiation between satellites. Each signal is pseudo randomly  generated and includes the time that it was transmitted along with some other information. Using 4 satellites it is possible to use the times generated by the satellites to accurately determine the amount of time the signal takes to reach the receiver. Using the time for the signal to reach the receiver and the known position of each satellite it is possible to say that the receiver is X distance from satellite 1, Y  distance from satellite 2 and Z  distance from satellite 3. Using the distances as a radius it is possible to find the location where each radius intersects. This intersection is the location of the receiver. Using a more accurate receiver that will pickup more satellites and letting it record for a longer time will produce a more accurate location for the intersection or radii. There are sources that create errors such as disturbances in the ionosphere, such as solar flares seen on earth as the Northern Lights.

For a more detailed description of how GPS works the following tutorials can be used.

• Produced by Trimble (Shockwave is a benefit)
• Satellite Navigation & Positioning group (SNAP) at The University of New South Wales (good technical discussion)
• By Javad and Nedda Ashjaee (A good history of location systems)

         Unfortunately there are times when not enough satellites are visible from a location. Earlier this year a project was done to rectify a photo of Thunder Bay using points located by GPS. For this project 10 points were taken, at each point 180 location were recorded spaced 1 second apart, to provide an accurate reading. Unfortunately there were not enough satellites around that were in receivable location so the project took around 6 hours to complete instead of 1 hour that it could have taken. This shows that it is not always easy to receive the signals from the satellites. A similar problem was encountered with this project, the comparison of GPS receivers. At times it was hard to receive signals from the satellites, and only 2 or 3 of the 4 needed could be received. Therefore some points were not taken, the points that were not taken can be seen in Table 2, all the no cells do not have any information but where the receiver was tried.

        The receivers were tried at four sites located around Thunder Bay, Ontario. The first site we went out to was at the end of Harbour View Express Way, in a black spruce (Picea Mariana) "swamp". The second was located at the Cascades Conservation Area in a Poplar (Populus sp.) stand, the third site was also located at the Cascades Conservation Area but in a Jack Pine (Pinus banksiana) stand. The forth site was located at a official benchmark located along Highway 11/17. Each of these sites had different challenges and issues that had to be overcome to get the GPS receivers to work well. At each of these sites a benchmark location was placed in using a high accuracy receiver that located the point to a within a centimetre. This allows for a good comparison of the locations each unit produced to a known control.

Site name	Picture of canopy at site	Site location	Best unit	Minimum Distance (metres)	Maximum Distance (metres)	Average Distance (metres)	GPS locations	Ranking
Spruce		Harbour View Expressway	Trimble Geo3 with Beacon	0.4	11.0	2.2		2
Poplar		Cascades Conservation area, end of Balsam St.	Sokkia Axis 3 with Beacon	1.0	17.1	4.7		3
Pine		Cascades Conservation area, end of Balsam St.	Sokkia Axis 3 with Beacon  and Trimble Geo XT with WAAS	1.4	17.1	5.8		4
Benchmark/ Monument		Highway 11/17 east of Balsam St.	Trimble PRoXRS with WAAS	0.4	5.0	2.0		1

        The Table below is a comparison of all the unit/correction type combinations. In the table the column called "best unit" was created by taking the average distance from benchmark and dividing it by the number of times the unit was tried minus the number of failed collections. The resultant number was used in the creation of the relative best unit value. These relative values were used to find the unit/correction type combinations that produced the most constant accurate results, these units are highlighted in yellow. After the best units were determined the other units were looked at. The green and blue lines were felt to be similar in comparison as they both had some errors with more errors going to the blue line, but the low average distance can balance these errors to some extent.

Looking at the data presented the best overall unit, independent of the correction type was the Geo XT produced by Trimble. The Geo XT was able to collect the first time at each of the four locations testing occurred at. Over the four locations it produced the lowest average distance to control point. So with a low error rate and a low average distance it was the best unit that we tested.

        With all the information collected for the comparison of GPS receiving units it is also possible to compare the type of correction used. The table below shows the resulting distances produced by each correction method, the percentage of errors and the most accurate method used. To find the best method the average distance was multiplied by one plus the error percent. The resulting values were then ranked in order of smallest to largest. The method that produced the best results was the beacon produced by Trimble. One problem with this method is that it requires a transmitting station located within 200 to 300 kilometres from where the GPS unit will be used.

        The testing of the GPS receivers had a number of inadequacies. Therefore if testing of GPS units was to occur again a number of changes should be made. The changes and reasoning behind them can be seen on this page.

Pictures taken while working on this project